Proposing a New PHP Fork

[yuxaoxao]

Hello EdmondDantes,

I am a PHP developer from China. I deeply regret the recent outcome of the official PHP proposal regarding true async functionality and find the conservative stance of the PHP community rather perplexing. In China, the internet industry is evolving rapidly, and we have seen the emergence of async solutions such as Swoole and Swow.

I would like to inquire if you would be interested in collaborating with proponents of PHP async programming to form a "PHP Async Working Group." The idea would be to fork a dedicated async-php-src for independent development. Admittedly, this is just a preliminary idea, and there may be numerous challenges to overcome. If you are open to this, perhaps we could begin by discussing the feasibility of establishing such a group with influential individuals and those possessing relevant technical expertise among the PHP async supporters.

Looking forward to your thoughts.

[supdger]

Although I cannot speak for everyone, time will prove that users of swoole and workerman support this approach.
The RFC for trueAsync has undergone multiple proposals and votes, garnering significant attention within the group. We express regret and indignation regarding the voting results for 1.5. The group unanimously supports yuxaoxao's idea, and we hope the ED will consider it.

[ds1109]

Swoole and Workerman users have shown clear support for this approach. After multiple proposals and significant engagement regarding the trueAsync RFC, we are disappointed by the 1.5 voting results. Given the group's unified support for yuxaoxao's idea, we hope the ED will review this decision.

[lly521]

Hi there,
I’m a PHP developer based in San Francisco, and I’ve been closely following the evolution of the language for years. Like many others in the community, I was genuinely disappointed by the recent outcome of the official PHP RFC proposing native asynchronous capabilities. What troubles me even more is the apparent reluctance within parts of the PHP community to embrace modern paradigms like true async I/O and cooperative concurrency—especially at a time when performance, scalability, and responsiveness are non-negotiable in real-world applications.
In San Francisco, the internet ecosystem moves fast. We’ve long outgrown the limitations of traditional request-per-process PHP models. That’s why projects like Swoole and Swow have gained such strong traction here—they’ve proven that PHP can absolutely thrive in high-concurrency, event-driven environments. These aren’t just experiments; they’re powering production systems at scale, serving millions of users with efficiency and reliability.
Given this momentum, I’d like to propose something bold: What if we formed an “Async PHP Working Group”—a coalition of developers, maintainers, and visionaries who believe PHP deserves first-class async support?
The idea isn’t to oppose the core PHP team, but to complement their work through focused, collaborative innovation. We could start by creating a dedicated space—perhaps even a purpose-built fork like async-php-src—where we prototype async-native extensions, refine coroutine models, and explore integration with modern event loops. Over time, this effort could generate the data, tooling, and community consensus needed to influence the official PHP roadmap.
Of course, this is still just a concept. There are technical, governance, and coordination hurdles ahead. But I truly believe that if we bring together key contributors—folks who’ve already built async solutions, framework authors open to new runtimes, and engineers running PHP in demanding environments—we could turn this vision into something tangible.
Would you be open to exploring this together? Even an initial conversation with a few like-minded developers could be the spark we need.
I’d love to hear your thoughts.
Best regards,

[EdmondDantes]

Hi everyone.

This repository is already an async fork of PHP. It has automated tests and project builds configured.
Although I haven't updated the master branch in a while, it's not a problem to do so and release PHP 8.5 Async + PHP 8.6 Async.

If you integrate the HTTP server from Swoole as a separate extension, you’ll get a fully complete solution.
This will make it possible to adapt Symfony and Laravel to run in a concurrent environment.

But all these efforts may end up being pointless without support from the PHP Foundation.

As for the working group creating one is a difficult task. You also need to understand the group’s purpose. What is it? To demonstrate that the current solution can work with a major framework? Yes, that’s probably achievable. But…

I don’t think that kind of proof will be enough. As one person wrote on INTERNALS: you can’t convince people if they don’t want to be convinced.

That’s why I believe the idea of creating such a group should come from the PHP Foundation and be supported by PHP itself. Only then will this work be effective.

[EdmondDantes]

To create a working group that can become a serious organization, you need:

1. At least one specialist from AMPHP
2. At least one specialist from ReactPHP
3. Representatives from the Swoole and Swow projects
4. Representatives of at least three or four major frameworks
5. FrankenPHP
6. And others...

Where the main goal is to demonstrate how the current approach makes it possible to develop applications using the existing infrastructure.

[supdger]

3. Representatives from the Swoole and Swow projects
   I think Swoole should not be a big problem. If other conditions are met, I don't think this will be an issue.
4. Representatives of at least three or four major frameworks
   There should be at least two or more that can respond.

[EdmondDantes]

3. Representatives from the Swoole and Swow projects
   I think Swoole should not be a big problem. If other conditions are met, I don't think this will be an issue.
4. Representatives of at least three or four major frameworks
   There should be at least two or more that can respond.

I can officially invite AMPHP members, and I think I can find ReactPHP developers as well.
I also know some of the Yii developers, and they claim that the new version can work in coroutines without changes. That would be interesting.

[jackmingren]

Hi EdmondDantes and everyone,

Just a regular PHP developer from China jumping in to say:

I strongly support this direction.
Even though I don’t have big resources or company backing, I represent thousands of ordinary developers like me who have been waiting for real async PHP for years.

We are tired of watching every true async RFC get rejected while Swoole/Swow prove every day that it works perfectly in production.

Please count my +1 and my moral support.
If all I can do right now is star the repo, spread the word, write blog posts, help translate docs, or test releases, I’m 100% in.

The Chinese PHP community is behind you.
Don’t give up.

Thank you for keeping this repository alive and for being willing to fight for it.